> I'm still wondering if HyperBros did ExecNG as a clean implementation
> of Exec (as MorphOS and Aros did), or they had access to OS sources
> and ExecNG wouldn't be created without it.

Given that the Friedens had access to the 3.1 sources (which surely nobody will deny), I wonder whether ExecSG could be called a proper clean-room reimplementation even in case they didn't really look at the 3.1 exec code. This is how it is typically done after all:

"Typically, a clean-room design is done by having someone examine the system to be reimplemented and having this person write a specification. [...] The specification is then implemented by a team with no connection to the original examiners."
https://en.wikipedia.org/wiki/Clean_room_design